Biological Assay of Vitamin E Application to Wheat Germ and Wheat Germ Oil L. S. PALMER, Division of Agricultural Biochemistry, University of Minnesota, University Farm, St. Paul, Minn.
E
’
certain particulars in which uniformity seems desirable. Certain details of procedure were adopted arbitrarily without supporting data to justify them, provided they seemed theoretically important and reasonable. The data secured were intended to determine primarily the “live implant efficiency” (L. I. E., percentage of placental implants which resulted in the birth of live young). The results, however, also made it possible to calculate other efficiency values-for example, a (T. I. E., “total implant efficiency”) value involving both live and still-born young. I n addition, littering efficiency values could be calculated, such as “live-litter efficiency” (L. L. E., percentage of pregnancies which result in litters containing living young) and “total litter efficiency” (T. L. E., percentage of pregnancies which result in live and still-born litters). The theoretical basis for recording individual still births and litters consisting exclusively of still-born young is the assumption that still-born young indicate some, although insufficient, vitamin E.
VANS and Burr (1) have adequately described the general procedure which must be followed in order to demonstrate vitamin E. Birth of young following administration of a known quantity of test substance to female rats of proved sterility because of vitamin E deficiency is the basis of Evans’ (3) rat unit of vitamin E . The smallest quantity of test substance which results in birth of young, when given in a single dose prior to the fifth day of gestation, is regarded as containing one rat unit. The a-tocopherol described by Evans, Emerson, and Emerson (2)?which produced this effect in 2.5-mg. dose, thus contained 400 rat units of vitamin E per gram of substance. Olcott and Mattill (8) and Olcott (7) had previously prepared concentrates of vitamin E having essentially the same potency. Suggestions have been made to reduce the size of the unit (and thus increase the number of calculated units per gram of substance) by dividing the effective single dose by the number of days of gestation in rats. Evans and Burr (1) have presented a few data in support of this by showing that a single dose of wheat germ oil has the same effect as when aliquots one twenty-first as great are given daily throughout gestation. However, t’le quantitative relationships thus indicated cannot be regarded as very convincing when it is considered that the quantity of vitamin E needed daily must increase greatly with advance in gestation if the vitamin is required primarily for cell nuclear function of the fetal and placental tissues, as suggested by the work of Mason (6) on testicular degeneration in E deficiency. Furthermore, it is difficult to visualize the processes by which oral doses of vitamin-containing product consumed during the last few days of gestation are digested and metabolized and the vitamin is effectively used by the cells which need it, when the animal is assumed to be receiving the minimum dose. Indeed, it has not yet been determined what is the latest day in the gestation period when a potent vitamin E product can be given by mouth with assurance that the vitamin will exert a physiological effect in that gestation. The conclusion seems justified that the single dose procedure or its essential equivalent is the only logical one to employ in a biological assay of vitamin E until more study has been given to questions raised in the foregoing discussion. A survey of the published vitamin E assays based on litters produced, from which one may calculate the units of E present per gram of test substance, shows only a general quantitative relation between different doses of the same substance and the percentage of animals littering. Data regarding the number of young born (both living and dead) do not improve the quantitative character of the test. I n view of the fact that vitamin E seems to function exclusively during gestation in connection with the nourishment of the fetus a t the site of implantation, observations regarding the percentage of implantations which give rise to living young in a properly conducted assay suggested possibilities for improving the expected quantitative relations between dose and effect. This was one of the principal reasons for the present study. Although the outcome was disappointing, the results have a certain qualitative value and may assist other workers to evaluate their results properly. At the outset it seemed desirable to bring together into one comprehensive statement the details of conducting a vitamin E assay. The published procedures lack standardization in
Procedure for Vitamin E Assay of Oils and Other Products BAXALRATION.:The following basal ration, which is a slight modification of the Evans (3) ration, is used throughout. Batches of the ration are kept in‘the refrigerator until used. Ingredient
% 27 36 22 10 4
at weekly intervals
2
PRELIMINARY PERIOD.Young female rats exactly 21 days old are placed in cages having raised screen floors and given the basal ration ad libitum. A number of rats of approximately the same age are kept in the same cage. At about 90 to 100 days of age, daily observations are begun of the sexual cycle, using the vaginal smear technic of Long and Evans (6). As soon as an approaching heat stage is noticed by the presence of large numbers of nucleated epithelial cells, the animal is mated with a normal male in a separate cage having a raised screen floor. If successful copulation is determined to have occurred through the presence of sperm and/ or plug in vagina or in tray of cage at the next daily examination, further examinations are omitted until the fifth and again on the tenth succeeding day. If no indication of sexual cycle is obtained it is assumed that the animal is pregnant. (Pseudo- regnancies occur rather frequently. In this case the animal wilrsometimes be in heat again on the ninth to eleventh day but more frequently on the twelfth to fourteenth day. In these cases the animal must be carried through another resorption before being used for assay.) The animal is kept isolated throughout in a single cage having a raised screen floor. Beginning the tenth day after impre’gnation the daily vaginal examination is resumed until the “erythrocyte sign” is secured, usually on the thirteenth to fifteenth day. (The “erythrocyte sign” is the term employed by Evans and Burr, 1, to “designate the appearance in the vaginal canal on the thirteenth, fourteehth. and fifteenth days of gestation of recognizable numbers of red blood corpuscles.”) If this is obtained, pregnancy is assured. Beginning the day the erythrocyte sign is obtained, the animal is weighed daily without further vaginal examination. If the weight increases and then decreases suddenly without evidence of any young on the twenty-first or twenty-second day and if manual examination at time of weighing also indicates the presence of fetuses which are never born or which disappear, or if weight remains essentially stationary (due to the presence of only one or two fetuses) without young being born, the conclusion is drawn that the animal is suffering from vitamin E deficiency. This 427
428
INDUSTRIAL AND ENGINEERING CHEMISTRY
VOL. 9, NO. 9
animal is now ready for use ploratory research or in the in a vitamin E assay. An EFFICIENCIES IN 54 VITAMIN E ASSAYS TABLE I. REPRODUCTION t e n t a t i v e assay of an unanimal which fails t o show Standard knom-at various doses it is resorption in the first gestaDeviation Kind of Efficiency \, Mean tion may be carried through a permissible to use three rats Total litter efficiencya 21.37 second gestation or even a Total implant effioiencyb 23.35 per dose, provided they are Live litter efficiencyc third, if necessary, to deplete 70.98 26.28 all successfully bred. It is Live i m d a n t efficiencvd \ 55.59 27.08 it of body stores of vitamin Efficiencies Correlated P. E. ( r ) expected that a t least two E. After a third failure to L. L. E. : T. I. E. $0.7935 0 034 show resorption the animal is out of the three animals in $0.6069 0.058 L. L. E.:L. I. E. discarded. $0.5430 0.065 T. L. E.: T. I. E. such a test will give posiCONDUCTING THE ASSAY. + O . 3206 0.082 T. L. E.:L. I. E. tive evidence of vitamin E Six female rats proved to be a Percentage of pregnancies resulting in live and still-born litters. potency in a potent test b Percentage of placental implantations resulting in live and still-born suffering from vitamin E deyoung. product. ficiency which have been kept c Percentage of pregnancies resulting in litters containing living young. d Percentage of placental implantations resulting in live young. continuously on the basal PERIOD OF R E T A I N I N G ration are examined daily for VITAMINE-DEFICIENT sexual cvcle a n d b r e d t o RATSBEFORE TEST. It is normal kales in s e p a r a t e cages at the next estrus. When impregnationis established as in assumed that a vitamin E-deficient animal is suitable for the preliminary period, the animals are isolated and kept in sepatest purposes any time within 3 months after its E deficiency rate cages. The vaginal examinations are omitted until the fifth has been established, provided it is kept on the basal ration. and tenth days, after which they are resumed daily until the QUALITY OF INGREDIENTS IN BASALRATION. The only inerythrocyte sign is obtained. Cages having raised screen floors are used throughout. gredient whose quality is of special importance is the cod liver In the meantime the standard or experimental dose of material oil, which must be U. S. P. grade. Each ingredient should repto be tested is administered to the animal in a single dose in a resent a single lot of substance for any given series of tests, separate nonscattering feed cup between the first and fourth day in order to avoid introduction of vitamin E into the basal of gestation, while withholding the basal ration. Oils are given, mixed carefully with sufficient dry yeast to make a friable paste, ration of an animal after establishing its deficiency and its or mixed with a few grams of basal ration from which the fats have use for assay purposes. been omitted; other solid material to be tested is offered in DISCUSSION.I n the author’s experience with assays weighed amount also in a separate nonscattering feed cup, the conducted according to this procedure, the chief cause of basal ration being withheld for that day. When the erythrocyte sign is obtained, the test animals are initial fertilities in female rats reared on the basal ration has weighed daily and the presence of fetuses is established as soon been starting the rats on the ration when more than 21 as possible by increase in live weight and by manual examination. days old. The most annoying difficulty has been a fairly Fetuses can usually be established by manual examination on the high percentage (10 to 20 per cent) of pseudo-pregnancies fourteenth day. Daily weighing is continued until parturition should occur. Animals which are about t o cast a litter, as shown after administration of test substance in spite of definitely by manual examination and weight increase, are placed on a successful matings with normal males. larger screen, 1-inch mesh, which permits the young to fall through to tray of cage. Resorptions can frequently be detected Experimental Results in advance by vaginal bleeding at some time subsequent to the erythrocyte sign and prior to the day of partbrition. However, The mean values, their standard deviation, and the corall rats in a test are carried through the twenty-second day of gestation, regardless of evidence of complete resorption of the relation coefficients between the different reproduction effifetuses. All animals are autopsied after parturition, or after the ciencies previously indicated are shown in Table I for 54 twenty-second day if no young are born, t o determine the number vitamin E assays in which 217 rats gave evidence of the of implantations and resorptions. vitamin E potency of 39 different samples of material of unknown vitamin E value. (The products tested were furnished ITERION OF VITAMINE POTENCY. If five of the six by General Mills, Inc., Research Laboratories, Minneapolis, animals become pregnant a t the mating immediately precedMinn.) ing the administration of the test substance and four of these The highest efficiency calculation is determined by countfive animals to which the standard or experimental dose of ing both live and dead litters in relation to the number of test substance has been administered give birth to normal pregnancies, the next highest when counting both live and living young corresponding to 10 per cent or more of the still births in relation to placental implantations, the third implantations, the vitamin E potency of the test substance is established. Record is kept of number and weight of young born, number of live and still births, and total number of implantations. TABLE11. RELATIONSHIPS BETWEEN RELATIVE DOSESOF VITAMINE AND RELATIVE EFFICIENCIES OF REPRODUCTION If an assay is started with a rat which proves not to have been pregnant a t the time the test material was administered Sample Total No. Relative Relative Relative Relative Relative Dose T . I . E . a T.L.E.b L.1.E.C L.L.E.d No. Rats Used the rat is continued in the assay if successfully bred a t the 75 150 1 6 240 280 150 next normal heat, but if this breeding fails the animal is dis19 67 67 e 2 19 678 6 33 72 e 67 3 6 58 33 carded. 4 6 67 e 67 108 105 67 e THESTANDARD DOSE OF TESTMATERIAL.This is the 100 5 89 82 6 73 100 100 41 59 50 100 4 6 minimum amount of test material administered in a single 100 83 108 50 7 6 33 82 67 8 20 64 a 100 73e dose which, when allowing for a margin of variation from 142 67 150 9 22 58 44 natural causes, will produce normal living young correspondMean 64.8 92 91.4 91 85.9 ing to 10 per cent or more of the implantations in the uterus Values Correlated r P.E. (T) Relative dose:relative T. I. E. , +0.455 0.178 of a vitamin E-deficient female rat and thus permit the stateRelative dose:relative L. L. E. +0.406 0.188 ment to be made that the product contains a measurable Relative dose:relative L. I. E. +0,321 0.202 Relative dose:relative T. L. E . 4-0.084 0.223 amount of vitamin E in such dose. USE OF LESSTHANFIVERATSPER TEST. The standard a Percentage of placental implantations resulting in live and still-born young. number of female rats to be placed on a single test for a b Percentage of pregnancies resulting in live and still-born litters. Percentage of implantations resulting in live young. standard aesay is six. Conclusions may be drawn if five of d Percentage of, pregnancies resulting in litters containing living young. the six animals are successfully bred a t the mating immediately * Relative efficiency within 10% of the relative dose. preceding the administration of the test substance. For ex-
SEPTEMBER 15, 1937
ANALYTICAL EDITION
429
highest when counting litters tive, primarily because the containing living young in T.4ELE 111. COMPARATIVE VITAMIN E ASSAYSOF WHE.4T GERM assay of the standard would AFD EXPELLER PROCESS:WHEATGERMOIL relation to pregnancies, and be subject to a large error. the 1ow e s t efficiency when No. of -E5cienoy ValuesThis may be too pessimistic T. L. E." T.I. E.b L. L. Sample Dose Rats L. I. E.d counting only living young a view. Grams % % % in relation to placental imThe data obtained in this Freah raw germ I 5.60 3 100 100 100 100 plantations. T h e m e t h o d study gave information with FreshoilfromgermI 0.75 3 100 88 100 75 Fresh raw germ I1 5.50 6 100 91 75 61 giving the highest values evireference to various vitamin Freah processed erm dently gives the most liberal (Embo) from% 6.50 6 100 100 E-containing products from 100 75 Freah processed germ interpretation of vitamin E wheat germ and may be as(Embo) I11 5.50 3 100 89 100 82 Stored rocesaed germ p o t e n c y and the m e t h o d sumed to have definite quali( E m h IV 5.50 6 100 78 83 56 giving the lowest values the tative value, at least. Table 5 Total litter efficiency. strictest interpretation. 111 shows comparative vitab Placental implant efficiency (live and etill-born). Live litter efficiency. Although there seems to be min E assays of raw wheat d Placental implant efficiency (live births). little choice between the plagerm and of the oil expelled c e n t al i m p l a n t efficiency from it, raw wheat germ and which considers both live and a processed w h e a t germ still births and the live litter efficiency, as judged by the mean (Embo) made from it, and fresh processed wheat germ and a values and standard deviations, the highest correlation cosample of the same kind of product after standing for one efficient is between the live litter efficiency and the live imyear a t room temperature. The samples in the latter complant &ciency. From a biological standpoint the latter corparison were taken for assay from a sealed evacuated tin relation seems reasonable even if the statistics do not, but in which the product is merchandised. biologically one would also expect a highly significant correlaIt is evident that the oil expelled from fresh raw wheat tion between the total litter efficiency (live and dead litters) germ had definite vitamin E potency, that the type of procand the total implant efficiency. Instead, this correlation essing employed in the case of the wheat germ samples tested coefficient is barely significant from a statistical standpoint. did not injure the potency, and that vitamin E was retained The probable reason is that too few rata were used in the sinin the processed germ to a considerable degree for a long period gle tests, the result being that a high percentage of the total in a sealed evacuated tin kept at room temperature. Thelatter litter efficiency values (63 per cent) were 100. Only oneconclusion is based on comparisons with the other samples of third of the living litter efficiency values were 100. processed germ tested, no assays having been made of the Although these results may give some basis for choosing original product when fresh. the most desirable data to secure in a vitamin E assay, they Table I V shows comparative vitamin E assays of several give no information as to the validity of such &ta in relation samples of wheat germ oil (expeller process) when fresh and to the quantity of vitamin administered. after storage a t refrigeration temDeratures. in sealed tin or glass contGiners. It Ts evident that the vitamin E was re, tained in these samples for a period of a t least 8 months in T m L B Iv. STABILITY OF VJTAMIN E IN EXPELLER-PROCEsS cold storage in sealed tin containers, and that no better presWHEATGERMOIL UNDER STORAGE ervation occurred in glass; only in sample C, stored for 12 Storage No. of -Efficiency Valuessample Age Conditions Dose Rats T. L. E.5 T. I. E.6 I;.L. E . c L. I. E . d was there evidence of loss of vitamin E. '
E . C
8
Milliurams % Sealed tjn 760 3 PO0 Sealed tin 760 6 83 1 Sealed tin 760 6 83 8 Sealed tin 760 6 100 1 Sealed tjn 500 3 67 C 12 Sealed tin 500 11 64 DI 5 Sealed tine 760 a 100 D: 5 Sealedglasd 750 3 100 0 Total litter efficienoy. b Placental implant efficiency (live and still-born) 0 Live litter efficiency. d Placental implant efficiency (li-ve births). 8 Kept at -20' F. Kept at 32' F. A A B
Mos.
%
%
%
3
88
100 83 67
75 72 47
67 27 67 67
17
6
83
65 64 17
$
28
83 92
83
48
8
79 67
I
~
Table I1 shows the results of the relation between each type of efficiency calculation and the dosage of vitamin obtained from a limited number of tests in which several products were administered at two different levels. The data shown are the relative levels and the corresponding relative efficiency values of each type, and the correlation coefficients. Possibly the samples are too few in number or an insufficient number of rats were employed in each test to warrant final conclusions; nevertheless the correlations are not significant. One is therefore forced to the conclusion that the assay of vitamin E thus conducted is at best merely a general quantitative assay and does not warrant very rigid interpretation in terms of units. Assuming that the relative efficiency should be within *10 per cent of the relative dose, it is seen that this occurred in only seven of the 36 cases, and only two of these occurred even when the number of rats employed was 20 or more. A reference standard, if available, probably would not make the results now obtainable more quantita-
Conclusions Biological assays of vitamin E, using oral administration to sterile, vitamin E-deficient rats in the earliest stages of pregnancy using a standardized procedure, have only a general quantitative value. However, results of such assays may be used for a comparison of potent vitamin E-containing products. When thus applied in this study to wheat germ and wheat germ oils it was found that there was (1) a close correlation between the vitamin E in the raw or processed wheat germ (Embo) tested and the pure oil expressed from the fresh germ; (2) a high retention of vitamin E in the processed (Embo) wheat germ kept for one year a t room temperature in a sealed, evacuated can; and (3) a high stability of vitamin E in the expressed wheat germ oil for several months a t refrigeration temperature in sealed containers, either glass or tin.
Literature Cited (1) Evans, H. M., a n d Burr, G. O., Mem. Univ. Calif., 8 (1927). (2) Evans, H. M., Emerson, 0. H., a n d Emerson, G . A,, J. Bid. Chem., 113, 319 (1936). (3) Evans, H. M., M u r p h y , E. A., Archibald, R. C.,a n d Cornish, R. E., Ibid., 108, 515 (1935). 14) H a w k , P. B., a n d Oser, B. L., Science, 74, 369 (1931). ( 5 ) Long, J. A . , a n d Evans, H. M., Mem. Unin. Calif., 6 (1922). (6) Mason, K. E., J. E x p t . Zodl., 45, 159 (1926). (7) Olcott, H. S., J. Bzol. Chem., 107, 471 (1934). (8) Olcott, H. S., a n d Mattill, H. A., J. Biol. Chem., 104, 423 (1934).
'
RECEIVBD May 10, 1937. Paper 1602, Scientific Journal Seriea, Minnesota Agricultural Experiment Station.
